StartUp

Inspiration Startups get one real shot to make an impression — at a hackathon, in front of an investor, or during a demo day. Many early-stage founders struggle not because their ideas are weak, but because they can't clearly articulate them. If you can't explain your problem, solution, and value proposition in 60 seconds, it's a signal that your business model may need refinement. We wanted to build the "HireVue for founders", a tool that helps entrepreneurs validate and sharpen their pitch before stepping into high-stakes rooms.

What It Does StartUp is an AI-powered pitch validation and investor-readiness platform. Founders record a 60-second pitch and receive an automatic transcription, followed by AI scoring across five dimensions: Clarity, Credibility, Investor Fit, The Ask, and Consistency. The platform provides structured scores out of 100, highlights various red flags, delivers actionable feedback, and tracks progress over time. GitHub repositories can also be integrated so founders can align their pitch with what they've actually built — ensuring they're not overpromising or underselling. Once a founder reaches a strong score of 80 or above, they unlock access to a curated investor directory and in-app messaging, encouraging meaningful improvement before outreach. StartUp turns an idea into a validated, investor-ready pitch.

How We Built It We built a full pitch evaluation pipeline consisting of several integrated layers. The recording layer uses the browser's MediaRecorder API to capture a 60-second video pitch. The audio or video file is then processed using OpenAI Whisper to generate a transcript. This transcript is sent to a large language model with a structured prompt that instructs it to evaluate the pitch like a venture capitalist, returning structured JSON scores and written feedback. Scores are saved to the user's profile to enable progress tracking over time. A GitHub integration layer pulls public repositories and compares claims in the pitch with actual implementation for alignment. The full system flow runs as follows: Video → Transcript → Structured AI Evaluation → Stored Scores → Actionable Insights.

Challenges We Ran Into Building StartUp came with a number of real challenges. Designing scoring criteria that felt fair and meaningful required different iterations. Engineering prompts that produced structured, consistent feedback was difficult, as was avoiding vague AI advice and ensuring suggestions were genuinely actionable. Integrating GitHub analysis meaningfully within hackathon time constraints pushed the limits of the team's bandwidth as well as balancing feature ambition with a clean and intuitive user experience was an ongoing tension throughout.

Accomplishments We're Proud Of The team built a fully functional end-to-end pitch scoring system with a structured evaluation of the startup. Progress tracking across multiple pitch attempts was implemented successfully, as was GitHub alignment for early product validation. Combining pitch coaching and investor outreach in a single platform — with a score-gated investor directory designed to promote improvement first — sets StartUp apart. More than a demo, the team delivered a working founder readiness system.

What We Learned Clarity is often the biggest weakness in early-stage pitches. When prompts are not precise, AI feedback becomes generic and unhelpful. Founders struggle most with defining their funding ask clearly, and a strong pitch does not guarantee a strong startup — but weak articulation is often a red flag. AI works best as a structured thinking partner rather than a final authority. Most importantly, the team learned that validation starts with articulation.

Built With

• github
• next.js
• openai